Logo Logo
 Search  |  Site Map  |  Contact Us  |  Login  |  Home 
Quick Search  Quick Search
About Agilent  |   Products & Services  |   Industries  |   International  |   Online Stores
Home > Products & Services > Test & Measurement > Technical Support > Metrology Forum
Metrology Forum : Basics

Measurement Procedure Adequacy

Introduction

Ensuring that your test equipment is properly and adequately calibrated isn't all that is required in order to make valid measurements. Limitations in the procedure defining how to make the measurement, including the equipment set-up conditions, can have a critical (and costly) impact !

A Simple Example

A process control equipment manufacturer chose to have their production test systems calibrated on-site to minimize downtime. During the calibration of one of the systems, an employee of the manufacturing company approached the calibration technician to express concern that the calibrations may have been done incorrectly. Apparently, ever since his test station had been calibrated, every product that was tested "Failed" and was rejected for rework

Slightly perplexed, the calibration technician returned to the suspect test station to investigate. The operator showed him the test procedure which stated that the ideal measurement result was 20, with limits of 19.90 to 20.10. He went on to explain that ever since the calibration that morning, each product he tested using the multimeter read "OVLD" (overload). This kind of problem hadn't happened when the former contractor calibrated the kit, he complained.

At once, the diligent technician asked about the ideal value. "Twenty what?", as no measurement units had been specified. Following further investigation, it was finally established that the test was actually to determine current in milli-amps, although the operator was unaware because the documented procedure failed to mention it. The multimeter had been correctly calibrated for all function and ranges -- but its function switch had been left in the position used for the last test... RESISTANCE !

A Complex Example : The 40 Million Dollar Mistake

The most well-known example of how a procedural mistake in testing can have a costly impact is the Hubble Space Telescope (HST). The story of what happened appears in the Appendix of our article Twenty-First Century Cal. which you can jump directly to below.

Explore

Articles

Basics

Contents

Download

Just for Fun

Metrology Forum Home

Metrology News & Events

Standards

Worldwide

Also in this Section
 
--------------------
Copyright 2000 Agilent Technologies | Terms and ConditionsPrivacy Statement